Human friendly teaching for industrial robots

Abstract

This paper considers a human friendly teaching method which can be used by every human being. The method represents robots' motion by a set of keywords which are frequently used in daily conversation. The sequence of keywords is interpreted as a sequence of robot command, and realized our intended motion. This paper also considers a method which translates human gesture to the keywords, because human gesture greatly assists to describe our intention as a set of keywords. To avoid mis-operation, we can also verify the consistency between the human gesture and typed keywords. Moreover, keyword description enables us to expand the interface to being more human friendly.